diff options
| author | Tuong Lien <tuong.t.lien@dektech.com.au> | 2020-05-26 12:38:36 +0300 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2020-05-27 01:16:52 +0300 |
| commit | a91d55d162b86fb983b88f44296149752db7efbd (patch) | |
| tree | f871946430b622fba0cb9acae24613138e78d700 /net/tipc/link.c | |
| parent | c6ed7a5cc2d68c36287c09260dc211173e0447d7 (diff) | |
| download | linux-a91d55d162b86fb983b88f44296149752db7efbd.tar.xz | |
tipc: enable broadcast retrans via unicast
In some environment, broadcast traffic is suppressed at high rate (i.e.
a kind of bandwidth limit setting). When it is applied, TIPC broadcast
can still run successfully. However, when it comes to a high load, some
packets will be dropped first and TIPC tries to retransmit them but the
packet retransmission is intentionally broadcast too, so making things
worse and not helpful at all.
This commit enables the broadcast retransmission via unicast which only
retransmits packets to the specific peer that has really reported a gap
i.e. not broadcasting to all nodes in the cluster, so will prevent from
being suppressed, and also reduce some overheads on the other peers due
to duplicates, finally improve the overall TIPC broadcast performance.
Note: the functionality can be turned on/off via the sysctl file:
echo 1 > /proc/sys/net/tipc/bc_retruni
echo 0 > /proc/sys/net/tipc/bc_retruni
Default is '0', i.e. the broadcast retransmission still works as usual.
Acked-by: Ying Xue <ying.xue@windriver.com>
Acked-by: Jon Maloy <jmaloy@redhat.com>
Signed-off-by: Tuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/tipc/link.c')
| -rw-r--r-- | net/tipc/link.c | 10 |
1 files changed, 6 insertions, 4 deletions
diff --git a/net/tipc/link.c b/net/tipc/link.c index 288c5670cfa5..af352391e2ab 100644 --- a/net/tipc/link.c +++ b/net/tipc/link.c @@ -375,7 +375,7 @@ void tipc_link_remove_bc_peer(struct tipc_link *snd_l, snd_l->ackers--; rcv_l->bc_peer_is_up = true; rcv_l->state = LINK_ESTABLISHED; - tipc_link_bc_ack_rcv(rcv_l, ack, 0, NULL, xmitq); + tipc_link_bc_ack_rcv(rcv_l, ack, 0, NULL, xmitq, NULL); trace_tipc_link_reset(rcv_l, TIPC_DUMP_ALL, "bclink removed!"); tipc_link_reset(rcv_l); rcv_l->state = LINK_RESET; @@ -2400,7 +2400,8 @@ int tipc_link_bc_sync_rcv(struct tipc_link *l, struct tipc_msg *hdr, int tipc_link_bc_ack_rcv(struct tipc_link *r, u16 acked, u16 gap, struct tipc_gap_ack_blks *ga, - struct sk_buff_head *xmitq) + struct sk_buff_head *xmitq, + struct sk_buff_head *retrq) { struct tipc_link *l = r->bc_sndlink; bool unused = false; @@ -2413,7 +2414,7 @@ int tipc_link_bc_ack_rcv(struct tipc_link *r, u16 acked, u16 gap, return 0; trace_tipc_link_bc_ack(r, acked, gap, &l->transmq); - tipc_link_advance_transmq(l, r, acked, gap, ga, xmitq, &unused, &rc); + tipc_link_advance_transmq(l, r, acked, gap, ga, retrq, &unused, &rc); tipc_link_advance_backlog(l, xmitq); if (unlikely(!skb_queue_empty(&l->wakeupq))) @@ -2447,7 +2448,8 @@ int tipc_link_bc_nack_rcv(struct tipc_link *l, struct sk_buff *skb, return 0; if (dnode == tipc_own_addr(l->net)) { - rc = tipc_link_bc_ack_rcv(l, acked, to - acked, NULL, xmitq); + rc = tipc_link_bc_ack_rcv(l, acked, to - acked, NULL, xmitq, + xmitq); l->stats.recv_nacks++; return rc; } |